Why High-Order Polynomials Should Not be Used in Regression Discontinuity Designs∗ Andrew Gelman† Guido Imbens‡ - Regression discontinuity design - Document - PDFSEARCH.IO

First Page	Meta Content
	Why High-Order Polynomials Should Not be Used in Regression Discontinuity Designs∗ Andrew Gelman† Guido Imbens‡ Add to Reading List Document Date: 2014-08-19 16:35:05 Open Document File Size: 414,42 KB Share Result on Facebook Company X5 / / Facility Columbia University / Stanford University / / Organization Graduate School / Institute for Education Sciences / National Science Foundation / Stanford University / NBER / Columbia University / New York / Department of Statistics / / Person Jennifer Hill / Van Der Klaauw / Joseph Cummins / Todd / / Position researcher / Porter / / Product pseudo / Pentax K-x Digital Camera / / ProvinceOrState New York / / SportsLeague Stanford University / / Technology simulation / / SocialTag Linear regression Local regression Regression discontinuity design Estimator Quadratic polynomial Errors and residuals in statistics Polynomial M-estimator Estimation theory Statistics

Why High-Order Polynomials Should Not be Used in Regression Discontinuity Designs∗ Andrew Gelman† Guido Imbens‡Add to Reading List